IntroductiontoDataMining.ppt

IntroductiontoDataMiningAnkurTeredesai,AssistantProfessor,,:AKDDProcessDatamining—coreofknowledgediscoveryprocessDataCleaningDataIntegrationDatabasesDataWarehouseKnowledgeTask-relevantDataSelectionDataMiningPatternEvaluationStepsofaKDDProcessLearningtheapplicationdomainrelevantpriorknowledgeandgoalsofapplicationCreatingatargetdataset:dataselectionDatacleaningandpreprocessing:(maytake60%ofeffort!)DatareductionandtransformationFindusefulfeatures,dimensionality/variablereduction,,classification,regression,association,(s)Datamining:searchforpatternsofinterestPatternevaluationandknowledgepresentationvisualization,transformation,removingredundantpatterns,,datawarehouses,andotherinformationrepositoriesWearedrowningindata,butstarvingforknowledge!Solution:DatawarehousinganddataminingDatawarehousingandon-lineanalyticalprocessingMininginterestingknowledge(rules,regularities,patterns,constraints)fromdatainlargedatabasesWhatIsDataMining?Datamining(knowledgediscoveryfromdata)Extractionofinteresting(non-trivial,implicit,previouslyunknownandpotentiallyuseful)patternsorknowledgefromhugeamountofdataAlternativenamesKnowledgediscovery(mining)indatabases(KDD),knowledgeextraction,data/patternanalysis,dataarcheology,datadredging,informationharvesting,businessintelligence,:Iseverything“datamining”?(Deductive):Whydatamining?Whatisdatamining?DataMining:Onwhatkindofdata?DataminingfunctionalityAreallthepatternsinteresting?ClassificationofdataminingsystemsMajorissuesindataminingAssumptionsDatabase:‘good’-–web?/Modeling/Classificat